NIH grant funds research to pinpoint natural selection’s influence on genomes

This post was originally published on this site

UNIVERSITY PARK, Pa. — The human genome — roughly 20,000 genes embedded in our DNA — is the result of millions of years of evolutionary forces as humans and humans’ ancestors struggled to survive. But scientists are unsure exactly how much influence natural selection, one of those evolutionary forces, has had on our genes. Now, through a $1.7 million grant through the National Institutes of Health, researchers led by Michael DeGiorgio, assistant professor of biology and Institute for CyberScience co-hire at Penn State, will begin to tease apart individual forces to understand how much influence natural selection has had on our evolutionary path.

DeGiorgio and his team will develop sophisticated statistical models that can help identify the various forces acting on human genes. Different versions of genes, known as mutations, are passed from generation to generation in a population for a variety of reasons, including randomness. One of these forces, natural selection, is the process of “choosing” genes based on their increased benefit to individuals in a specific environment. For example, around many speedy predators, the combination of genes that helps make prey fast is more likely to survive naturally.

DeGiorgio will use statistical learning to figure out which forces in the past contributed to the current set of genes and genetic mutations that remain in humans today. Their approach is analogous to taking a meal in a restaurant and using your knowledge of cooking processes and the appearance of individual molecules on your plate to determine whether the food was cooked in a frying pan or in the oven, and at what temperature. In DeGiorgio’s team’s case, they will be pinpointing whether mutations were new or had existed for generations, how quickly beneficial mutations spread through populations, and other important biological criteria.

It’s a complicated challenge because little is known about the environmental conditions of our early human ancestors. DeGiorgio and other computational biologists have shown that supercomputers can run through thousands of possible scenarios until they find one that matches patterns observed in humans’ genes today. DeGiorgio’s work will use the Institute for CyberScience Advanced CyberInfrastructure, Penn State’s high-performance research cluster.

“It’s only because we have sophisticated computing now that we can do thousands or hundreds of thousands of simulations to actually get this distribution of genetic diversity we should expect under different processes,” DeGiorgio said.

Their work will improve the understanding of which genes are important because of natural selection, which could have medical implications, said DeGiorgio. For example, in past work, DeGiorgio compared the genetic influences of Ethiopians living at high altitudes to those at low altitudes.

“These studies identified genes that are highly relevant medically; for example, genes involved in hypoxia response at high altitudes where there are low oxygen levels,” he said.